Jökull - 01.12.1965, Blaðsíða 11

Fig. 8. Section through the Grímsvötn atea following the lines A — B — C — D on the map.
Dash line shows the lower limit of the ice after a jökulhlaup, assuming that the ice is floating,
dot line the caklera floor on the same premises. Filled circles: Goocl or fair soundings, open
circles: poor soundings. Circle with x inside: Sounding by A. Joset 1951. — Snið gegnum Grims-
vatnasvceðið fylgjancli linunum A — B — C — D á 7. mynd. Strikalina sýnir neðri mörk iss, ef
hann er fljótandi, punkialina boin öskjunnar, ef svo er. Fylllir hringir tákna öruggar eða scemi-
legar rncelingar, tómir hringir óábyggilegar mœlingar. Hringur um x er samkvcemt mcelingu A.
Joset 1951.
detect any ice-water contact on his íilms. This
does not necessarily mean that tliere was in
reality no water in the caldera. A water-ice
contact may be very difficult to detect and as
a matter of fact there niust ltave been a lot of
water tliere at least in 1951, as the last pre-
ceding jökulhlaup occurred three years earlier,
in February 1948, and we know with certainty
that water increases continuously and at rather
even a rate in the caldera between jökulhlaups
because of ablation as well as continuous sub-
glacial melting 'by solfatara activity and heat
outflow.
But if there really was water underneath the
ice when the soundings were carried out then
the calculated thickness figures must be re-
duced as they were based on seismic wave velo-
cities in ice and not in water, and if the ice
is floating the ltedrock floor of the caldera is
at the depth indicated bv the dot-line on the
section Fig. 8. If, on the other hand, we accept
the possibility that the thickness of the ice is
such or nearly such as Holtzscherer has cal-
culatecl from the seismic soundings, which
means tliat all water stored in the caldera is
drained off during the jökulhlaups, then the
ice cannot be floating at all, not even imme-
diately before a hlaup. The flatness of the firn
surface in the central caldera, even immediately
after a jökulhlaup, must then be due to a
similar flatness of the bedrock floor of the
caldera — such a flatness we have in the Askja
caklera. The lake at the foot of Vatnshamar—
Depill is then á lateral lake the level of which
is not inl'luenced by the water stored on the
bottom of the caldera. But then it is difficult
to explain why the height difference between
tlie lake surface and the firn surface remains
so remarkably constant.
The release mechanism for the Grímsvötn
jökulhlaups will be discussed further in a later
paper dealing with the accumulation and lrud-
get of the Grímsvötn area. The former opin-
ion of the present writer (Thorarinsson 1953 a)
that it is a lifting of the damming icebarrier
which trigs off the hlaup, seerns doubtíul in
JÖKULL 117